The Role of the Automotive Industry in Detroit's Employment Fluctuations: A Multiple Restriction Regime Approach

Using a vector autoregressive time-series model, I analyze the validity of a multiple restriction regime approach in regional modeling as well as the historical impact of various employment shocks on the Detroit Metropolitan Area economy. For this, the model's decomposition matrix has two distinct forms that are dependent on the relation of an industry to the automotive industry, the region's largest manufacturing sector, and the causal restrictions these different regimes place on each version of the model. Results indicate that the use of two different sets of restrictions yields acceptable explanatory power and captures important dynamic characteristics of the data without a substantial increase in computational costs over a single restriction regime approach. Results also indicate that declines in employment in the automotive industry over the past two decades did have a negative effect on the Detroit Metropolitan Area, but not of the immediacy or magnitude popularly believed.